Tanguy De Keyzer

De Keyzer sees marketing caught between two separate but equally urgent shifts: the automation of repetitive work, and the fundamental rewiring of how people discover information.

On automation, he is pragmatic and specific about scope. AI should handle "all repetitive" tasks—data movement, report generation, competitor tracking, podcast repurposing —but only to free humans for strategy, creativity, and customer understanding . He dismisses the either-or framing entirely. The constraint he identifies is not capability but role clarity: "The art sits in the division of labor, not in the choice" . He sees AI winning on "speed and volume," humans winning on "experience, opinion, and original insights" . He warns explicitly against treating content production as fully automatable; 70% of purchasing decisions rest on emotional factors, and AI-generated material at scale produces "generic text" that erodes this advantage .

The deeper shift concerns how visibility itself works. He argues that ranking in Google is becoming "pointless if AI doesn't know you" . Search is fragmenting—ChatGPT, Claude, Gemini, and Perplexity each have different architectures, indexes, and selection behaviors —and no single optimization approach transfers across them . The metric that matters is not clicks but citation frequency within AI knowledge networks . Being "named" in an AI response, he argues, carries more weight than a top search ranking because it functions as repeated expert validation across thousands of daily conversations .

His answer to this reordering is what he calls "becoming source material." This requires, in sequence: making genuine value that deserves visibility, ensuring that value is structurally extractable by models, and building real authority through consistent quality . He is direct about the prerequisite: "You cannot optimize an empty shell into a source" . The competitive advantage belongs not to those who game the algorithm most cleverly but to those who become genuinely indispensable—a position he notes "takes competitors years to replicate" .